adask-b/.claude/skills/k8s-platform-expert

Kubernetes Platform Expert

A comprehensive Kubernetes skill combining deployment expertise with systematic troubleshooting capabilities. Covers the full lifecycle from design and deployment through incident response and remediation.

When to Use This Skill

Deployment & Configuration:

• Deploying workloads (Deployments, StatefulSets, DaemonSets, Jobs)
• Configuring networking (Services, Ingress, NetworkPolicies)
• Managing configuration (ConfigMaps, Secrets)
• Setting up persistent storage (PV, PVC, StorageClasses)
• Creating and managing Helm charts
• Implementing security best practices (RBAC, PSS)

Troubleshooting & Incident Response:

• Investigating pod failures (CrashLoopBackOff, ImagePullBackOff, Pending)
• Responding to production incidents or outages
• Diagnosing cluster health issues
• Troubleshooting networking or storage problems
• Debugging ArgoCD sync failures
• Troubleshooting Helm release problems
• Diagnosing kind cluster issues (local development)

---

Part 1: Deployment & Security

Core Deployment Workflow

1. Analyze requirements - workload characteristics, scaling needs, security requirements
2. Design architecture - workload types, networking patterns, storage solutions
3. Implement manifests - declarative YAML with proper resource limits, health checks
4. Secure - apply RBAC, NetworkPolicies, Pod Security Standards
5. Test & validate - verify deployments, test failure scenarios

Deployment Constraints (MUST DO)

• Use declarative YAML manifests (avoid imperative kubectl commands)
• Set resource requests AND limits on all containers
• Include liveness and readiness probes
• Use Secrets for sensitive data (never hardcode)
• Apply least privilege RBAC permissions
• Implement NetworkPolicies for network segmentation
• Use namespaces for logical isolation
• Label resources consistently
• Pin image versions (never use latest in production)

Deployment Constraints (MUST NOT DO)

• Deploy without resource limits
• Store secrets in ConfigMaps or plain env vars
• Use default ServiceAccount for application pods
• Allow unrestricted network access (default allow-all)
• Run containers as root without justification
• Skip health checks
• Expose unnecessary ports or services

Output Templates

When implementing Kubernetes resources, provide:

• Complete YAML manifests with proper structure
• RBAC configuration (ServiceAccount, Role, RoleBinding)
• NetworkPolicy for network isolation
• Brief explanation of design decisions

---

Part 2: Troubleshooting & Incident Response

Core Troubleshooting Workflow

1. Gather Context

   • What is the observed symptom?
   • When did it start?
   • What changed recently?
   • What is the scope (pod, service, node, cluster)?
   • What is the business impact?

2. Initial Triage Run cluster health check:

   python3 .claude/skills/k8s-platform-expert/scripts/cluster_health.py
   

3. Deep Dive Investigation

   Namespace-Level:

   python3 .claude/skills/k8s-platform-expert/scripts/check_namespace.py <namespace>
   

   Pod-Level:

   python3 .claude/skills/k8s-platform-expert/scripts/diagnose_pod.py <namespace> <pod-name>
   

   ArgoCD Issues:

   python3 .claude/skills/k8s-platform-expert/scripts/diagnose_argocd.py <app-name>
   

   Helm Issues:

   python3 .claude/skills/k8s-platform-expert/scripts/diagnose_helm.py <release> <namespace>
   

4. Identify Root Cause - consult references/common-issues.md

5. Apply Remediation - test in non-prod first, document actions, have rollback ready

6. Verify & Monitor - confirm fix, monitor 15-30 min minimum

Incident Severity Levels

| Level | Description | Examples | |-------|-------------|----------| | SEV-1 | Critical | Complete outage, data loss, security breach | | SEV-2 | High | Major degradation, significant user impact | | SEV-3 | Medium | Minor impairment, workaround available | | SEV-4 | Low | Cosmetic, minimal impact |

Quick Reference Commands

Cluster Overview:

kubectl cluster-info
kubectl get nodes
kubectl get pods --all-namespaces | grep -v Running
kubectl get events --all-namespaces --sort-by='.lastTimestamp' | tail -20

Pod Diagnostics:

kubectl describe pod <pod> -n <namespace>
kubectl logs <pod> -n <namespace>
kubectl logs <pod> -n <namespace> --previous
kubectl exec -it <pod> -n <namespace> -- /bin/sh

Node Diagnostics:

kubectl describe node <node>
kubectl top nodes
kubectl top pods --all-namespaces

Service & Network:

kubectl describe svc <service> -n <namespace>
kubectl get endpoints <service> -n <namespace>
kubectl get networkpolicies -n <namespace>

Storage:

kubectl get pvc,pv -n <namespace>
kubectl describe pvc <pvc> -n <namespace>
kubectl get storageclass

---

Best Practices

Always:

• Start with high-level health check before deep diving
• Document symptoms and findings
• Check recent changes (deployments, config, infrastructure)
• Preserve logs before making destructive changes
• Test fixes in non-production when possible
• Monitor after applying fixes

Never:

• Make production changes without understanding impact
• Delete resources without confirming safety
• Restart pods repeatedly without investigating root cause
• Apply fixes without documentation
• Skip post-incident review

---

Related References

• references/common-issues.md - Detailed troubleshooting guides
• references/deployment-patterns.md - Workload patterns
• references/security-hardening.md - Security best practices

---

Sources

Combined from:

• Jeffallan/claude-skills - kubernetes-specialist
• ahmedasmar/devops-claude-skills - k8s-troubleshooter